arX

iv:1

008.

3736

v2 [

phys

ics.

ins-

det]

24

Aug

201

0

arXiv: 1008.3736 [physics.ins-det] 26 April 2010

Gaseous Detectors: recent developments and

applicationsa

Maxim TitovCEA Saclay, DSM/IRFU/SPP, 91191 Gif sur Yvette, France

Since long time, the compelling scientific goals of future high energy physics experiments werea driving factor in the development of advanced detector technologies. A true innovationin detector instrumentation concepts came in 1968, with the development of a fully parallelreadout for a large array of sensing elements - the Multiwire Proportional Chamber (MWPC),which earned Georges Charpak a Nobel prize in physics in 1992. Since that time radiationdetection and imaging with fast gaseous detectors, capable of economically covering large de-tection volume with low mass budget, have been playing an important role in many fields ofphysics. Advances in photo-lithography and micro-processing techniques in the chip industryduring the past decade triggered a major transition in the field of gas detectors from wirestructures to Micro-Pattern Gas Detector (MPGD) concepts, revolutionizing cell size limita-tions for many gas detector applications. The high radiation resistance and excellent spatialand time resolution make them an invaluable tool to confront future detector challenges atthe next generation of colliders. The design of the new micro-pattern devices appears suitablefor industrial production. Novel structures where MPGDs are directly coupled to the CMOSpixel readout represent an exciting field allowing timing and charge measurements as well asprecise spatial information in 3D. Originally developed for the high energy physics, MPGDapplications has expanded to nuclear physics, UV and visible photon detection, astroparticleand neutrino physics, neutron detection and medical physics.

1 Gaseous Detectors: Historical Overview

The single wire proportional counter, invented more 100 years ago by E. Rutherford and H.Geiger1, and its high gain successor, the Geiger-Mueller counter first described in 19282, can beconsidered the ancestors of all modern gaseous detectors and were for many decades a major toolfor the study of ionizing radiation. Forty years ago, in 1968, the instrumentation in experimentalparticle physics was revolutionized with the advent of the Multi-Wire Proportional Chamber(MWPC) 3. With its excellent accuracy and modest rate capability, the MWPC allowed largedetector areas to be instrumented with fast tracking detectors and were able to localize particletrajectories with sub-mm precision. Confronted by the increasing demands of particle physics

aInvited Lectures at the Trans-European School of High Energy Physics, Zakopane, Poland, July 8-14 (2009);http://events.lal.in2p3.fr/TESchool09/

1

experiments, MWPCs have continuously improved over the years 4,5,6,7,8. Gradually replacingslower detectors, numerous generations of gaseous devices, with novel geometries and exploit-ing various gas properties, has been developed: Drift Chamber 9, Multi-Drift Module 10, JETChamber 11, Time Projection Chamber (TPC) 12, Time Expansion Chamber 13, Multi-Step

Chamber 14, Ring-Imaging Cherenkov Counter 15 and many others. However, limitations havebeen reached in maximum rate capability and detector granularity. A fundamental rate limita-tion of wire chambers, due to positive ion accumulations, was overcome with the invention ofMicro-Strip Gas Chamber (MSGC)16, capable of achieving position resolution of few tens of mi-

crons at particle fluxes exceeding MHz/mm2 17. Developed for the projects at high-luminositycolliders, MSGC’s filled a gap between the performing but expensive solid-strip detectors andcheap but rate-limited traditional wire chambers. Despite their impressive performance, detailedstudies of long-term behavior at high rates have revealed two possible weaknesses of the MSGCtechnology: a slow degradation under sustained irradiation (aging), and appearance of rare but

destructive discharges in presence of highly ionizing particles 18.

The invention of Micro-Pattern Gas Detectors (MPGD), in particular the Gas Electron Mul-

tiplier (GEM)19, the Micro-Mesh Gaseous Structure (Micromegas) 20, and other micro patterndetector schemes, offers the potential to develop new gaseous detectors with unprecedentedspatial resolution, high rate capability, large sensitive area and operational stability 21,22. Re-cent developments in radiation hardness research with state-of-the-art MPGD’s revealed thatthey might be even less vulnerable to radiation-induced performance degradation than stan-dard silicon micro-strip detectors23,24. In some applications, requiring very large-area coveragewith moderate spatial resolutions, more coarse macro-patterned detectors, e.g. thick-GEMs(THGEM) 25,26,27 or patterned resistive thick GEM devices (RETGEM) 28 could offer an in-teresting and economic solution. In addition, the availability of highly integrated amplificationand readout electronics allows for the design of gas-detector systems with channel densities com-parable to that of modern silicon detectors. An elegant solution is the use of a CMOS pixelASICb, assembled directly below the GEM or Micromegas amplification structures 29,30,31,32.Modern wafer post-processing also allows for the integration of Micromegas grid directly ontop of a CMOS pixelized readout chip 33. In 2008, the RD51 collaboration at CERN has beenestablished to further advance technological developments of micro-pattern gaseous detectorsand associated electronic-readout systems, for applications in basic and applied research 34.

2 Basic Principles: Ionization, Transport Phenomena and Avalanche Multiplica-tion

The process of detection in gas proportional counters starts with the inelastic collisions betweenthe incident particles and gas molecules. These collisions lead to excitation of the medium(followed by the emission of the light, the basis of scintillation detectors) and ionization, theprimary signal for tracking devices. The number NP and the space distribution of the primaryionization clusters depend on the nature and energy of the radiation. The primary electronscan often have enough energy to further ionize the medium; the total number of electron-ionpairs (NT ) is usually a few times larger than the number of primaries (NP ). Table 1 providesvalues of relevant parameters in some commonly used gases at NTP (normal temperature and

pressure) for unit charge minimum-ionizing particles (MIPs)35. Values often differ, dependingon the source, and those in the table should be taken only as approximate.

The primary statistics determines several intrinsic performances of gas detectors, such asefficiency, time resolution and localization accuracy. The actual number of primary interactionsfollows the Poisson’s statistics; the inefficiency of a perfect detector with a thin layer of gas is

bApplication Specific Integrated Circuit

2

Table 1: Properties of noble and molecular gases at normal temperature and pressure (NTP: 20 C, one atm).EX , EI : first excitation, ionization energy; WI : average energy per ion pair; dE/dx|min, NP , NT : differentialenergy loss, primary and total number of electron-ion pairs per cm, for unit charge minimum ionizing particles.

Gas Density Ex EI WI dE/dx|min NP NT

mg cm−3 eV eV eV keV cm−1 cm−1 cm−1

He 0.179 19.8 24.6 41.3 0.32 3.5 8

Ne 0.839 16.7 21.6 37 1.45 13 40

Ar 1.66 11.6 15.7 26 2.53 25 97

Xe 5.495 8.4 12.1 22 6.87 41 312

CH4 0.667 8.8 12.6 30 1.61 28 54

C2H6 1.26 8.2 11.5 26 2.91 48 112

iC4H10 2.49 6.5 10.6 26 5.67 90 220

CO2 1.84 7.0 13.8 34 3.35 35 100

CF4 3.78 10.0 16.0 54 6.38 63 120

given by exp(-NP ). For example, in one mm of Ar/CO2 (70:30) approximately 6 % of all MIPsdo not release a single primary electron cluster and therefore can not be detected. The totalenergy loss, sum of primary and secondary ionization, follows a statistical distribution describedby a Landau function, with characteristic tails towards higher values. A simple compositionlaw can be used for gas mixtures: for example, the number of primary (NP ) and total (NT )electron-ion pairs produced by MIP in a 1 cm of Ar/CO2 (70:30) mixture at NTP:

NP = 25 · 0.7 + 35 · 0.3 = 28pairs

cm; NT =

2530

26· 0.7 + 3350

35· 0.3 ≈ 97

pairs

cm(1)

While charged particles release ionization trail of primary electron clusters, low energy X-rays undergo a single localized interaction, usually followed by the emission of the photo-electron,accompanied by the lower-energy photon or Auger electron. For example, a 5.9 keV X-rayconverts in argon mainly on a K shell (3.2 keV ); the emitted photo-electron with energy Eγ −EK ∼ 2.7 keV has a practical range in detector of ∼ 200 µm. In addition, with 85 % probabilityanother (Auger) electron with energy ∼ 3.2 keV (∼ 250 µm range in argon) is ejected; in theremaining cases, a 3 keV K-L fluorescence photon is produced with a mean absorption length of40 mm. The sum of the energies of photo-electron and Auger electron is responsible for the main5.9 keV peak, while fluorescence mechanism leads to the Ar escape peak. The total number ofelectron-ion pairs created by X-ray absorbed in argon can be evaluated by dividing its energyby the WI :

5900

26≈ 227.

Once released in the gas, and under the influence of an applied electric field, electrons andions drift in opposite directions and diffuse towards the electrodes. When electron move throughthe gas, the magnitude of the scattering cross section in an electron-atom (molecule) collision isdetermined by the details of atomic and molecular structure. Therefore, the drift velocity anddiffusion of electrons depend very strongly on the nature of the gas, namely on the structure ofinelastic cross-section, arising from rotational and vibrational levels of molecules. In noble gases,the inelastic cross section is zero below excitation and ionization thresholds. Large drift velocitiesare achieved by adding polyatomic gases (usually CH4, CO2 or CF4), having large inelastic crosssections at moderate energies, which results in “cooling” electrons into an energy range of theRamsauer-Townsend minimum (located at ∼0.5 eV) of the momentum transfer elastic cross-section of argon. The reduction in both the total electron scattering cross-section and theelectron energy results in a large increase of electron drift velocity (for a compilation of electron-

molecule cross sections see Ref. 36). Another principal role of the polyatomic gas is to absorb

3

the ultraviolet (UV) photons emitted by the excited inert gas atoms. The quenching of UVphotons occurs through the photo-decomposition of polyatomic molecules. Extensive collectionsof experimental data37 and theoretical calculations based on transport theory38 permit estimatesof drift and diffusion properties in pure gases and their mixtures. In a simple approximation, gaskinetic theory provides the following relation between drift velocity, v, and the mean collisiontime between electron and molecules, τ (Townsend’s expression): v = eEτ/m. Values of driftvelocity and diffusion for some commonly used gases at NTP are given in Fig. 1a and Fig. 1b.These have been computed with the MAGBOLTZ program39. Using fast CF4-based mixtures atfields around kV/cm−1, the electron drift velocity is around 10 cm ·µs−1 and the correspondingdrift time 100 ns·cm−1. For different conditions, the horizontal axis must be scaled inversely withthe gas pressure or density, 1/P , where P is the pressure. Standard deviations for longitudinal(σL) and transverse diffusion (σT ) are given for one cm of drift, and scale with the square rootof distance. Since the collection time is inversely proportional to the drift velocity, diffusion issmaller in gases having large drift velocities, such as for CF4. In the presence of an externalmagnetic field, the Lorentz force acting on electrons between collisions deflects the driftingelectrons and modifies the drift properties. For parallel electric and magnetic fields, drift velocityand longitudinal diffusion are not affected, while the transverse diffusion can be strongly reduced:σT (B) = σT (B = 0)/

√1 + ω2τ2. The dotted line in Fig. 1b represents σT for the classic Ar/CH4

(90:10) mixture at 4 T. This reduction is exploited in TPC to improve spatial resolution.

a) b)

Figure 1: a) Computed electron drift velocity with the MAGBOLTZ program as a function of electric field inseveral gases at NTP and B=0. b) Electron longitudinal diffusion (σL) (dashed lines) and transverse diffusion(σT ) (full lines) for 1 cm of drift at NTP and B = 0. The dotted line shows σT for the Ar/CH4(90:10) gas at 4T.

In mixtures containing electronegative molecules such as O2, H2O or CF4, electrons canbe captured to form negative ions. Capture cross-sections are strongly energy-dependent, andtherefore the capture probability is a function of applied field. For example, the electron isattached to the oxygen molecule at energies below 1 eV, while dissociative electron attachmentto the CF4 occurs mainly in the 6 to 8 eV range 40. The three-body electron attachmentcoefficients may differ greatly for the same addition in different mixtures. As an example, atmoderate fields (up to 1 kV/cm) the addition of 0.1% of oxygen to an Ar/CO2 mixture resultsin an electron capture probability about twenty times larger than the same addition to Ar/CH4.

The primary ionization signal is very feeble in a gas layer: in one cm of Ar/CO2 (70:30) atNTP around ∼ 100 electron-ion pairs are created (see Eq. 1). Therefore, one has to use “internalgas amplification” mechanism to generate detectable signal in gas counters; excitation and subse-quent photon emission participate in the avalanche spread processes and can be also detected byoptical means. If the electric field is increased sufficiently, electrons gain enough energy betweencollisions to undergo inelastic collisions with gas molecules. Above a gas-dependent threshold,the mean free path for ionization, λi, decreases exponentially with the field; its inverse, α = 1/λi,

4

is the first Townsend coefficient. In single wire proportional counter, most of the increase ofavalanche particles occurs very close to the anode wires, and a simple electrostatic considera-tion shows that the largest fraction of the detected signal is due to the motion of positive ionsreceding from the wires. The electron component, although very fast, contributes very little tothe signal. This determines the characteristic shape of the detected signals in the proportionalmode: a fast rise followed by a gradual increase. The slow component, the so-called ”ion tail”that limits the time resolution of the counter, is usually removed by differentiation of the signal.In uniform fields, N0 initial electrons multiply over a length x forming an electron avalanche ofsize N = N0 e

αx; N/N0 is the gain of the counter. With present-day electronics, proportionalgains around 5 · 103 − 104 are sufficient for detection of minimum ionizing particles, and noblegases with moderate amounts of polyatomic gases, such as methane or carbon dioxide, are used.

Positive ions released by the primary ionization or produced in the avalanches drift anddiffuse under the influence of the electric field. Negative ions may also be produced by electronattachment to gas molecules. The drift velocity of ions in the fields encountered in gaseouscounters (up to few kV/cm) is typically about three orders of magnitude lower than for electrons.The ion mobility, µ, the ratio of drift velocity to electric field, is constant for a given ion type upto very high fields 41,42. For mixtures, due to a very effective charge transfer mechanism, onlyions with the lowest ionization potential survive after a short path in the gas. The diffusion ofions, both σL and σT , are proportional to the square root of the drift time, with a coefficientthat depends on temperature but not on the ion mass. Accumulation of ions in the gas volumemay induce gain reduction and field distortions.

3 The Multi-Wire Proportional, Drift and Time Projection Chambers

The invention of the Multi-Wire Proportional Chambers (MWPC) revolutionized the field of

radiation detectors 3,5. In the original design, the MWPC consists of a set of parallel, evenlyspaced, anode wires stretched between two cathode planes. Applying potential difference be-tween anodes and cathodes, field lines and equipotentials develop as shown in Fig. 2a. Electronsreleased in the gas volume drift towards the anodes and produce avalanche in the increasingfield. Typical values for the anode wire spacing range between 1 and 5 mm, the anode to cath-ode distance is 5 to 10 mm. The operation gets increasingly difficult at smaller wire spacings,which prevented taking this direction for obtaining higher spatial resolution. For example, theelectrostatic repulsion for thin (10 µm) anode wires causes mechanical instability above a crit-ical wire length, which is less than 25 cm for 1-mm wire spacings. The signal multiplicationprocess, which begins a few radii from the anode, is over after a fraction of nanosecond, leavingthe cloud of positive ions receding from the anode. Detection of charge on the anode wiresover a predefined threshold provides the transverse coordinate to the wire with an accuracycomparable to that of the wire spacing. With a digital readout and s = 1 mm wire spacing,the spatial resolution is limited to: σ = s√

12= 300 µm. The coordinate along each wire can

be obtained by measuring the ratio of collected charge at the two ends of resistive wires. Theslow motion of positive ions produced in the avalanche is responsible for the largest fractionof charge signals detected on all surrounding electrodes; a measurement of the charge profileinduced on segmented cathodes, allows bi-dimensional localization of the ionizing event. This isthe so-called center-of-gravity (COG) method, which permits to attain the highest localizationaccuracy in MWPCs. Due to the statistics of energy loss and asymmetric ionization clusters,the position accuracy is ∼ 50 µm rms for tracks perpendicular to the wire plane, but degradesto ∼ 250 µm at 30 to the normal 43.

Drift chambers, developed in the early ’70’s, can estimate the position of a track by exploitingthe arrival time of electrons at the anodes if the time of interaction is known9. The distancebetween anode wires is usually several cm allowing coverage of large areas at reduced cost. Many

5

−0.20

−0.15

−0.10

−0.05

0.00

0.05

0.10

0.15

x-axis [cm] −0.2−0.3 −0.1 0.0 0.1 0.2 0.3

−0.15

−0.10

−0.05

0.00

0.05

0.10

0.15

0.20

y-ax

is [

cm]

y-ax

is [

cm]

(a) Multiwire proportional chamber

(b) Drift chamber

Figure 2: Electric field lines and equipotentials in (a) a multiwire proportional chamber and (b) a drift chamber.Thin anodes wires between two cathodes acts as a set of independent proportional counters.

designs have been introduced aimed at improving performance. In the original design, a thickerwire at proper voltage between anodes (field wire) reduces the field at the middle point betweenanodes, improving charge collection (Fig. 2b). In some drift chambers design, and with the helpof suitable voltages applied to field-shaping electrodes, the electric field structure is adjustedto improve the linearity of space-to-drift-time relation, resulting in better spatial resolution 44.Drift chambers can reach a spatial resolution from timing measurement of order 100 µm (rms)or better for minimum ionizing particles, depending on geometry and operating conditions. Adegradation of resolution is observed however for tracks close to the anode wires, caused by thespread in arrival time of the nearest ionization clusters, due to primary ionization statistics 45.Sampling the drift time on rows of anodes led to the concept of multiple arrays such as themulti-drift module 10 and the JET chamber 11. A measurement of drift time, together withthe recording of charge sharing from the two ends of the anode wires provides the coordinatesof segments of tracks; the total charge gives information on the differential energy loss and isexploited for particle identification.

The “ultimate” drift chamber is the TPC concept invented in the 1976 12, which combinesa measurement of drift time and charge induction on cathodes to obtain excellent tracking forhigh multiplicity topologies occurring at moderate rates. It has been the prime choice for largetracking systems in e+e− colliders (PEP-4 46, ALEPH 47, DELPHI 48) and proved its unique

resolving power in the heavy ion collisions (NA49 49, STAR 50). A TPC consists of a largegas volume, with an uniform electric field applied between the central electrode and a grid at

6

the opposite side. The ionization trails produced by charged particles drift towards the readoutend-plate, where a 2D image of tracks is reconstructed; the third coordinate is measured usingthe drift time information. In all cases, a good knowledge of electron drift velocity and diffusionproperties is required. This has to be combined with the 3D modeling of electric fields inthe structures, computed with commercial or custom-developed software 39,51. Conventionalreadout structure, based on MWPC and pads, is a benchmark for the “most modern” ALICETPC, designed to cope with extreme instantaneous particle densities produced in heavy ioncollisions at the LHC. This detector incorporates innovative and state of the art technologies,from the mechanical structures to the readout electronics and data processing chain 52. Tolimit distortions to the ALICE TPC intrinsic spatial resolution (σrφ ∼1000 µm for 250 cm drift

length), the temperature gradient in a 88 m3 gas volume space must not exceed 0.1C 53. For

an overview of detectors exploiting the drift time for coordinate measurement see Ref. 7 andRef. 8.

Despite various improvements, position-sensitive detectors based on wire structures are lim-ited by basic diffusion processes and space charge effects in the gas to localization accuraciesaround 100 µm54. The presence of slow moving positive ions from electron avalanches generatesa positive space charge in the drift tube, that modifies electric field and leads to the uncertaintyin the space-to-drift time relation. In standard operating conditions, the gain of a MWPC startsto drop at particle rates above 104 mm−2s−1, leading to a loss of detection efficiency 44 (seeFig. 3). Together with the practical difficulty to manufacture detectors with sub-mm wire spac-ing, this has motivated the development of new generation gaseous detectors for high luminosityaccelerators.

Figure 3: Normalized gas gain as a function of particle rate for Multi-Wire Proportional Chamber (MWPC) andMicro-Strip Gas Chamber (MSGC).

4 Micro-Pattern Gas Detectors (MPGD)

Modern photo-lithographic technology led to the development of novel Micro-Pattern Gas Detec-tor (MPGD) concepts 5, revolutionizing cell size limitations for many gas detector applications.By using pitch size of a few hundred microns, an order of magnitude improvement in granularityover wire chambers, these detectors offer intrinsic high rate capability (> 106 Hz/mm2), excellentspatial resolution (∼ 30 µm), multi-particle resolution (∼ 500 µm), and single photo-electrontime resolution in the ns range.

The Micro-Strip Gas Chamber (MSGC), a concept invented in 1988, was the first of the

micro-structure gas detectors16. It consists of a set of tiny parallel metal strips laid on a thinresistive support, alternatively connected as anodes and cathodes (see Fig. 4a). The principle ofMSGC resembles a multi-anode proportional counter, with fine printed strips instead of wires,

7

see Fig. 4b. Owing to the small anode-to-cathode distance (∼ 100 µm), the fast collection of pos-itive ions by nearby cathode strips reduces space charge build-up, and pushes the maximum ratecapability of MSGC, compared to MWPC, by more than two orders of magnitude17 (see Fig. 3).Despite their promising performance, experience with MSGCs has raised serious concerns abouttheir long-term behavior. There are several major processes, particularly at high rates, leadingto the MSGC operating instabilities: substrate charging-up and time-dependent modification ofelectric field, surface deposition of polymers (aging) and destructive micro-discharges under ex-

posure to heavily ionizing particles 5,55. The problem of discharges is the intrinsic limitation ofall single-stage micro-pattern detectors in hadronic beams 56,57,58. Whenever the total chargein the avalanche exceeds a value of 107−108 electron-ion pairs (Raether limit), an enhancementof the electric field in front of and behind the primary avalanche induces the fast growth of afilament-like streamer followed by breakdown. This has been confirmed under a wide range ofoperating conditions and multiplying gaps 59,60,61,62. In the high fields and narrow gaps, theMSGC turned out to be prone to irreversible discharges induced by heavily ionizing particlesand destroying the fragile electrode structure18 (see Fig. 4c). Nevertheless, the detailed studieson their properties, and in particular, on the radiation-induced processes leading to dischargebreakdown, led to the development of mature technologies and novel approaches (GEM and Mi-cromegas) with similar performances, improved reliability and radiation hardness. The MPGDstructures can be grouped in two large families: micromesh-based detectors and hole-type struc-tures. The micromesh-based structures include: Micromegas, “Bulk” Micromegas, “Microbulk”Micromegas and “InGrid”. The Hole-type structures are: GEMs, THGEM, RETGEM andMicro-Hole and Strip Plate (MHSP) elements.

a) b) c)

Figure 4: a) Schematic view, equipotentials and field lines in the MSGC; b) Microscopic image of the MSGC. Onan insulating substrate, thin metallic anode strips alternate with wider cathodes; the pitch is 200 µm; c) Image ofMSGC electrodes damaged by discharge. The very thin metal layers of MSGCs (few hundred nanometers) makes

them vulnerable for discharges, which can easily destroy fragile structure.

140 µm

50 µm

a) b) c)

Figure 5: a) Schematic view and typical dimensions of the hole structure in the GEM amplification cell. Electricfield lines (solid) and equipotentials (dashed) are shown; b) Electron microscope view of a GEM foil etched on acopper-clad, 50 µm thick polymer foil. The hole’s diameter and pitch are 70 µm and 140 µm; c) Schematic view

of the triple-GEM detector.

8

Introduced in 1996 by F. Sauli 19, a Gas Electron Multiplier (GEM) detector consists ofa thin-foil copper-insulator-copper sandwich chemically perforated to obtain a high density ofholes. The GEM manufacturing method, developed at CERN, is a refinement of the double-side printed circuit technology. The hole diameter is typically between 25 µm and 150 µm,while the corresponding distance between holes varies between 50 µm and 200 µm. The centralinsulator is usually (in original design) the polymer kapton, with a thickness of 50 µm (seeFig. 5b). Controlled etching of GEM foils (decreasing the thickness of the copper layer from 5to 1 µm) allows to reduce material budget in triple-GEM to 1.5×10−3X0, which is about one

half of a 300-µm-thick Si-microstrip detector 63. Application of a potential difference betweenthe two sides of the GEM generates the electric fields indicated in Fig. 5a. Each hole acts as anindependent proportional counter. Electrons released by the primary ionization particle in theupper conversion region (above the GEM foil) drift into the holes, where charge multiplicationoccurs in the high electric field (50 - 70 kV/cm). Most of avalanche electrons are transferred

into the gap below the GEM 64,65. Several GEM foils can be cascaded (see Fig. 5c), allowingthe multi-layer GEM detectors to operate at overall gas gain above 104 in the presence of highlyionizing particles, while strongly reducing the risk of discharges (< 10−12 per hadron) 66,67.This is a major advantage of the GEM technology. A unique property of GEM detector is a fulldecoupling of the amplification stage (GEM) and the readout electrode (PCB), which operate atunity gain and serves only as a charge collector. Localization can then be performed by collectingthe charge on a one- or two-dimensional readout board of arbitrary pattern, placed below thelast GEM. The signal detected on the PCB is entirely due to the electron collection, without aslow ion tail, and is typically few tens of nanosecond for 1 mm-wide induction gap. CascadedGEMs reach gains above 105 with single electrons; this permitted conceiving gaseous imagingphoto-multipliers (GPM) with single photon sensitivity68. Moreover, with an appropriate choiceof GEM fields and geometry, both photon and ion feedback can be considerably suppressed.

The MHSP is a GEM-like hole-electrode with anode- and cathode-strips etched on the bot-tom face of the GEM; the avalanche developed inside the hole is further multiplied on the anodestrips 69. When operating the MHSP and biasing the strips in reverse mode, i.e. using an extraelectrode to attract positive ions, a breakthrough in ion blocking capability has been achievedin MHSP/GEM cascaded multipliers 70.

HV1

HV2Micromesh

100 µ

m

Anode plane

e−

E2

40 kV/cm

Particle

Drift gap

Amplificationgap

a) b) c)

Figure 6: a) Schematic view of the Micromegas detector (not to scale). A few mm drift gap is between thecathode electrode and micromesh, and the 100 µm amplification gap is defined by the micromesh and the anodestrip readout; b) Electric field map in Micromegas. A metallic micromesh separates a low-field (∼ 1 kV/cm) driftregion from the high-field multiplication region (∼ 40 − 80 kV/cm). c) Photograph of the “Bulk” Micromegas

detectors. Pillars of 400 µm diameter every 2 mm are visible.

Introduced in 1996 by Y. Giomataris, the Micromegas is a thin parallel-plate avalanchecounter, as shown in Fig. 6a20. It consists of a few mm drift region and a narrow multiplicationgap (25-150 µm) between a thin metal grid (micromesh) and the readout electrode (strips orpads of conductor printed on an insulator board). To preserve a distance between the anode andthe micromesh, regularly spaced pillars from insulating material are used. Electrons from the

9

primary ionization drift through the holes of the mesh into the narrow multiplication gap, wherethey are amplified. The electric field is homogeneous both in the drift (electric field ∼1 kV/cm)and amplification (∼40-80 kV/cm) gaps (see Fig. 6b). Due to the narrow multiplication region,small variations of the amplification gap are counteracted by an inverse variation of the Townsendcoefficient, thus improving the uniformity and stability of response over a large area71. Positiveions are quickly removed by the micromesh; this prevents space-charge accumulation and inducesvery fast signals with only a small ion tail, 50 - 100 ns length. The Micromegas retains the ratecapability and energy resolution of the parallel-plate counter. The small amplification gap is akey element in Micromegas operation, giving rise to excellent spatial resolution: 12 µm accuracy,limited by the micromesh pitch, has been achieved for MIPs 72, a very good energy resolution(∼11% FWHM with 8 keV X-rays) 73 and single photo-electron time resolution better than

1 ns74. Efforts have been focused on producing the Micromegas amplification region as a singlepiece using the newly developed “Bulk” method75. A woven mesh is laminated on a PCB coveredby a photo-imageable polyimide film, and the pillars are made by a photochemical techniquewith insulation through the grid. Such a “all-in-one” detector, called “Bulk” Micromegas, isrobust and allows the regular production of large, stable and unexpensive detector modules.Fig. 6c shows photo of the large “Bulk” Micromegas (36*34 cm2 active area) produced for the

T2K TPC76. A new Micromegas manufacturing technique, based on kapton etching technology,has been recently developed, resulting in further improvement of the detector characteristics,such as flexible structure, low mass, high radio-putiry, uniformity and stability 77. Excellentenergy resolution has been obtained, reaching 11 % FWHM for the 5.9 keV X-rays and 1.8 %FWHM of the 5.5 MeV α-peak for the Am241 source.

The success of GEMs and glass capillary plates triggered the development of coarse and morerobust structures, “optimized GEM” 25,78 followed by THGEM 26,27,79 gaseous multiplier.These are produced by standard printed circuit technology: mechanical drilling of 0.3-1 mmdiameter holes, etched at their rims to enhance high-voltage stability (see Fig. 7a and 7b);different PCB materials can be used, of typical thicknesses of 0.4-1 mm and hole spacing of0.7-1.2 mm. These electron multipliers exhibit specific features: the geometrical parameterscan be scaled from GEM ones, but the microscopic behavior of the electrons, in particulardiffusion in the gas, does not scale. The electron collection and transport between cascadedelements is more effective than in GEM because the THGEM’s hole diameter is larger than theelectron’s diffusion range when approaching the hole. The THGEM detectors have been studiedin laboratory, also in gaseous-photomultiplier configurations - coupled to semitransparent orreflecting CsI photo-cathodes (see Fig. 7a) 80. Effective gas amplification factors of 105 and 107

can be reached in single and cascaded double-THGEM elements, which permits efficient imagingof light at single-photon level79. Stable operation at photon fluxes exceeding 1 MHz/mm2 was

recorded together with sub-mm localization accuracy and timing in the 10 ns range 27,81.

a) b) c)

Figure 7: a) Schematic view of a double-THGEM with a reflective CsI PC deposited on top of first THGEM; b)Photo of the Thick GEM (THGEM) multiplier. A rim of 0.1 mm is chemically etched around the mechanically

drilled holes to reduce discharges; c) Photo of the RETGEM detector with resistive kapton electrodes.

10

A novel spark-protected version of thick GEM with resistive electrodes (RETGEM) has beenrecently developed, where the Cu-clad conductive electrodes are replaced by resistive materi-als 82,83,84. Sheets of carbon loaded kapton 50 µm thick or screen-printed resistive surfaceare attached onto both surfaces of the PCB to form resistive-electrode structure; holes 0.3 mmin diameter with a pitch of 0.6 mm are mechanically drilled (see Fig. 7b). At low countingrates, the detector operates as a conventional THGEM with metallic electrodes, while at highintensities and in case of discharges the behavior is similar to a resistive-plate chamber.

5 Micro-pattern Gas Detector Applications

The performance and robustness of MPGDs have encouraged their use in high-energy and nuclearphysics, UV and visible photon detection, astroparticle and neutrino physics, X-ray imaging andneutron detection and medical physics. Common themes for future applications are low mass,large active areas, high spatial resolution, high rate capabilities and high radiation tolerance.

Due to the wide variety of geometries and flexible operating parameters, MPGDs are acommon choice for tracking and triggering detectors in nuclear- and particle-physics. COMPASS,a high-luminosity experiment at CERN, pioneered the use of large-area (∼ 40 × 40 cm2) GEMand Micromegas detectors close to the beam line with particle rates of 25 kHz/mm2. Bothtechnologies achieved a tracking efficiency of close to 100% at gas gains of about 104, a spatialresolution of 70–100 µm and a time resolution of ∼ 10 ns 85,86. For the future COMPASSphysics program, a set of Micromegas detectors with pixelized readout in the central region isbeing proposed 87. High resolution planar triple-GEM detectors are used in the LHCb MuonSystem88, for the TOTEMTelescopes89 and also being developed for PANDA experiment at thefuture FAIR facility90, SBS spectrometer for Hall A at Jefferson lab91, and as a forward Trackerof STAR experiment at RHIC 92. Using fast CF4-based mixtures, a time resolution of about5 ns rms is achieved 93, adequate to resolve two bunch crossings at the Large Hadron Collider(LHC). Therefore, large area MPGDs are currently being considered for the LHC machineupgrade program (sLHC), aiming to increase the luminosity by a factor of ∼ 10, as a potentialreplacement of conventional gaseous detectors. Micromegas coupled to the CMOS readout maybe employed as new vertex detector system at the sLHC94. GEMs and Micromegas can be alsobent to form cylindrically curved ultra-light detectors, as preferred for inner tracker (barrel) and

vertex applications 95,96.

For the future International Linear Collider applications, both GEM and Micromegas de-vices are foreseen as one of the main options for the TPC97,98,99. Compared to wire chambers,they offer number of advantages: negligible E×B track distortion effects, the narrow pad re-sponse function (PRF) and the intrinsic suppression of ion back-flow, relaxing the requirementson gating of the devices and, depending on the design, possibly allowing non-gated operation ofTPCs. Large-area MPGD systems are also being studied as a potential solution for highly gran-ular digital hadron calorimeter. Implementations in GEM 100, Micromegas 101, and THGEMtechnologies have been proposed.

In the recent years there has been a considerable progress in the field of gaseous photo-multipliers by combining MPGD with semi-transparent or reflective CsI-photocathodes (PC)

to localize single photoelectrons 102,103,104. The operation of MPGD-based photomultipliersin CF4 with CsI-PC could form the basis of new generation windowless Cherenkov detectors,where both the radiator and the photosensor operate in the same gas. Exploiting this princi-ple, originally proposed for Parallel Plate Avalanche Chamber 105, a Hadron Blind Detectorwas successfully operated using triple-GEM amplification system with CsI-PC for the PHENIXexperiment at RHIC 106,107. A hadron-blindness property is achieved by reversing the direc-tion of the drift field ED, therefore pushing primary ionization produced by charged particlestowards the mesh. In this configuration photoelectrons released from the CsI-PC surface are

11

still effectively collected into the GEM holes due to the strong electric field inside holes. Formany applications in photon detectors, a rather coarse (sub-mm) spatial resolution is usuallysufficient. Therefore THGEM and RETGEM are under study for RICH detector upgrades forCOMPASS and ALICE108.

a) b)

Figure 8: a) Operation principle of triple-GEM with CsI-PC in the Hadron Blind Detector; b) Operation principleof the two-phase avalanche detector with a triple-GEM and reflective CsI-PC; both ionization and scintillation

signals from the liquid are detected.

Recent progress in the operation of cascaded GEM, Micromegas, THGEM and RETGEMGPMs at cryogenic temperatures (down to 80 K) 109 and in two-phase mode 110,111,112 couldpave the road towards their potential applications for the next generation neutrino physics andproton decay experiments 113,114, direct dark matter searches 115, Positron Emission Tomog-raphy (PET) 116, and for noble-liquid Compton telescope, combined with a micro-PET cam-

era 117,118. The operation principle of the cryogenic two-phase avalanche detector with GEMreadout is demonstarted in Fig. 8b: the ionization produced in the noble liquid by radiationis extracted from the liquid into the gas phase by an electric field. The multi-GEM detector,operated at cryogenic temperature in saturated vapor above the liquid phase, can detect boththe ionization signal, extracted from the liquid, and the scintillation signal, generated in thenoble liquid by a particle 111. The latter is achieved by depositing an UV-sensitive photocath-ode, namely CsI, on top of the first GEM (see Fig. 8b). The detection of both scintillation andionization signals could allows efficient background rejection in rare-event experiments; in PETapplications, the detection of scintillation signals could provide a fast trigger for coincidencesbetween two collinear gamma-quanta.

X-ray counting and imaging detectors, based on MPGDs, are being used for diffractionexperiments at synchrotron radiation facilities 119 and could serve as a powerful diagnostictool for magnetic fusion plasmas 120. An innovative GEM-based system, which combines fastimaging of X-ray emissions with spectral resolution in the VUV range (0.2-10 keV ), has beendeveloped to study 2D dynamics of plasma instabilities and to control core plasma position, bothbeing crucial issues for fusion researches120. Another recent development is the construction ofthe spherical GEM detector for parallax-free X-ray diffraction measurements, produced by theCERN/RD51 group in collaboration with industry 121. In the rare-event experiments, MPGDsare used in searching for solar axions (CAST), where the expected signal comes from solar axionsconversions into low-energy photons of 1-10 keV range. Micromegas with high granularity anodeelements can largely reduce the background event rate down to 5×10−5keV−1cm−2s−1, exploitingits stability, good energy and spatial resolution 122.

There are many applications of the Micromegas in the neutron detection domain, whichinclude neutron beam diagnostics 123, inertial fusion experiments 124, thermal neutron tomog-raphy 125 and a sealed Picollo-Micromegas detector, designed to provide in-core measurementsof the neutron flux and energy (from thermal to several MeV) in the nuclear reactor 126. Neu-trons can be converted into charged particles to detect ionization in Micromegas by two means:

12

either using the detector gas filling or target with appropriate deposition on its entrance window.

There were attempts to use GEM detectors for medical physics and portal imaging. In par-ticular, GEM-based prototype was used to detect simultaneously the position of the therapeuticradiation beam (γ’s) and the position of the patient tumor, using X-rays, in order to providefeedback to the cancer treatment machine and to correct on-line position of the beam with re-spect to the patient 127,128. In this device, the X-ray image was obtained using conversionsin gas and the γ-beam profile was determined using solid converters placed in between severalGEM foils. Another medical application is the radiation therapy, which demands new on-linebeam monitoring systems with ∼ 1 mm spatial resolution and 3D dosimetry of delivered doseswith an accuracy of 5 %. A scintillation light produced in the double-GEM detector duringavalanche development was detected by commercial CCD camera. With a 360 MeV α-beamthe scintillating GEM light signal at the Bragg peak was only 4 % smaller than of the referenceionization chamber. Consequently, GEM detector with CCD readout could become a feasiblesubstitute for the Lanex scintillating screen, especially for α- or carbon-ion beams.

Finally, MPGDs can be used for a variety of security related applications: detection of dan-gerous cargo 130, radon detection in the air as an early warning of earthquakes 131 and UVsensitive early forest fire detection system 133,132. Muon tomography, based on the measure-ment of multiple scattering of cosmic ray muons traversing cargo of vehicles that contain high-Zmaterials, is a promising passive interrogation technique to detect hidden nuclear materials.Large area GEM detectors for precise tracking of atmospheric muons in combination with af-fordable electronics and readout system is being developed by one of the RD51 collaboratinginstitutes 130. Among the planetary disasters the most common and often happening are theearthquakes and forest fires. A sharp increase in the Rn concentration before earthquakes hasbeen observed in the air regions associate to rocks and caves and its detection can serve as abasis of an early earthquake warning system. The RETGEM-based device, capable to operatein the air and to continuously monitor the Rn concentration, is currently under development.A network of such detectors can installed in the sensitive regions to provide daily assessment ofthe Rn concentration131. The same RD51 group is also developing a RETGEM-based detector,which could be used for detection of forest fires at distances up to 1 km, compared with a rangeof 200 m for commercially available UV-flame devices 133,132.

6 Development of large area MPGDs

A big step in the direction of the industrial manufacturing of large-size MPGD’s with size of fewm2 and spatial resolution typical of silicon micro-strip devices (30− 50 µm) is the development

of the new fabrication technologies - single mask GEM 134 and “Bulk” Micromegas 75. Meth-ods of their effective production and characterization are under investigation within the RD51collaboration.

Recent developments on large area GEMs are focused on two new techniques to overcomethe existing limitations: a single-mask technology and a splicing method for GEM foils 134.The standard technique for creating the GEM hole pattern, involving accurate alignment oftwo masks, is replaced by a single-mask technology to pattern only the top copper layer. Thebottom copper layer is etched after the polyamide, using the holes in the polyamide as a mask.A single mask technique overcomes the cumburstone practice of alignment of two masks betweentop and bottom films within 5-10 µm, which limits the achievable lateral size to ∼ 50 cm. In asplicing procedure, foils are glued over a narrow seam, obtaining a larger foil. Both techniqueswere successfully implemented in the large-area prototype of 66 × 66 cm2 size, produced incollaboration with CERN TS-DEM workshop, shown in Fig. 9a. Single mask GEM seems to bemuch more suitable for industrial large scale production than standard GEM technology.

The basic idea of the “Bulk” Micromegas technology is to built the whole detector in a single

13

process: the anode plane with copper strips, a photo-imageable polyamid film and the wovenmesh are laminated together at a high temperature forming a single object 75. At the end,the micromesh is sandwiched between two layers of insulating material, which is removed afterUV exposure and chemical development. This “Bulk” Micromegas technique has been recentlyextended to produce large area prototypes in CERN TS-DEM workshop, up to 150× 50 cm2 ina single piece (see Fig. 9b) 135.

The THGEM and RETGEM technologies could offer economically interesting solution fora single-photon Cherenkov imaging counters: namely good spatial and time resolutions, largegains (single photo-electron sensitivity), relatively low mass and easy construction - thanks tothe intrinsic robustness of the PCB electrodes. To advance future developments of GPMs, oneof the key ingredients is the possibility to produce, with good quality and yield, THGEM-likefoils of large surfaces coated with CsI UV-sensitive photocathodes. There is an ongoing R&Dwithin the RD51 collaboration to optimize THGEM geometry and to study charging up effects,gain stability and maximum achievable gain (similar to GEMs). Fig. 9c shows a first 60×60 cm2

THGEM foil produced in cooperation with industry (ELTOS, Italy) 136.

a) b) c)

Figure 9: a) The triple-GEM prototype of 66 × 66 cm2 active area, using single-mask GEM technology, forTOTEM experiment; b) Large area “Bulk” Micromegas prototype of 40 × 130 cm2 size for the sLHC ATLAS

muon system upgrade; c) First 60× 60 cm2 THGEM foil produced for the COMPASS RICH upgrade.

The development of large-area position-sensitive photon detectors, with visible sensitivephotocathodes (e.g. bialkali), could lead to numerous spin-off, beyond the Cherenkov lightimagers. Most commonly-used in the visible range are vacuum photomultipliers (PMTs), withrather limited module size and bulky geometry due to mechanical constraints on the glass vacuumenvelope. A possible alternative is to use gas-filled photomultipliers at atmospheric pressure;a proof of principle of visible-sensitive GPM was demonstrated with MPGDs 103,137. Futurestudies with large-area detection and visible-sensitive PC are being planned within the RD51collaboration.

7 Pixel readout for micro-pattern gas detectors

Coupling of the micro-electronics industry together with advanced PCB technology has beenvery important for the development of modern gas detectors with increasingly smaller pitchsize. The fine granularity and high-rate capability of micro-pattern devices can be furtherexploited by introducing high-density pixel readout with a size corresponding to the intrinsicwidth of the detected avalanche charge. However, for a pixel pitch of the order of 100 µm,technological constraints severely limit the maximum number of channels that can be broughtto the external front-end electronics. While the standard approach to readout MPGDs is a

14

segmented strip or pad plane with front-end electronics attached through connectors from thebackside, an attractive possibility is to use CMOS pixel chip, assembled directly below the GEMor Micromegas amplification structures, and serving as integrated device hosting the input pad,the preamplification and the digitization of the signals 29,30,31,32. Due to the strong electricfield in the GEM or Micromegas amplification structures, primary electron initiates avalanche,and the resulting charge signal activates the preamp-shaper-discriminator circuitry, integratedin the underlying active layers of the CMOS technology. Using this approach, gas detectorscan reach the level of integration, compactness and resolving power typical of solid-state pixeldevices.

a) b)

Figure 10: a) Photo of the analog CMOS ASIC with hexagonal pixels, bonded to the ceramic package; b) Photoof the Medipix2 chip; a 25 µm wide conductive bump bond openings, used for electron collection, are seen as a

matrix of dots.

The combination of GEM detector and an analog, low-noise and high granularity (50 µmpitch) CMOS pixel ASIC, comprising pixellated charge collecting anode and readout electron-ics, can bring large improvement in sensitivity, at least 2 orders of magnitude, compared totraditional X-ray polarimeters (based on Bragg diffraction or Compton scattering) 139,140,141.Moreover, scattering polarimeters are practically insensitive below 5 keV and are backgroundlimited, while Bragg crystal polarimeters are efficient only around a narrow band fulfilling theBragg condition. In contrary, the pixel readout of MPGDs allows to reconstruct individuallow-energy (2-3 keV) photo-electron tracks with a length as short as few hundred microns; thetotal charge collected in the pixels is proportional to the X-ray energy. The high detectorgranularity allows to recognize the initial part of the track, before it is distorted by Coulombscattering, through the localization of absorption point of the photon and then to estimatethe photo-electron emission direction. Finally, the degree of X-ray polarization is computedfrom the distribution of reconstructed track angles, since the photo-electron is emitted mainlyin the direction of photon electric field. Three ASIC generations of increased complexity andsize, reduced pitch and improved functionality have been designed and built 30,142,143,144 (seeFig.10a). The third ASIC version, realized in 0.18 µm CMOS technology, includes self-triggeringcapability and has 105600 hexagonal pixels with a 50 µm pitch, corresponding to an active areaof 15×15 mm2. Single GEM detector coupled to such a CMOS pixel chip can convert photonsin the energy range from a few keV up to tens of keV, by choosing the appropriate gas mixture,and is able to simultaneously produce high resolution images (50 µm), moderate spectroscopy(15% FWHM at 6 keV) and fast timing (30 ns) signals. At the focal plane of the large areamirror like XEUS telescope, the novel device will allow to perform energy-resolved polarimetryat the level of few % on many galactic and extragalactic sources with photon fluxes down to onemilliCrab in one day 143,144.

The original motivation of combining a MPGD with Medipix2 145 and Timepix 146 chipswas the development of a new readout system for a large TPC at the future Linear Collider.The digital Medipix2 chip was originally designed for single photon counting by means of asemiconductor X-ray sensor coupled to the chip. In gas detector applications, the chip is placed

15

in the gas volume without any semiconductor sensor, with GEM or Micromegas amplificationstructure above it 31,147,32. Approximately 75 % of every pixel in the Medipix2 matrix iscovered with an insulating passivation layer. Hence, the avalanche electrons are collected onthe metalized input pads, exposed to the gas (see Fig. 10b). The Timepix chip, which is amodification of the Medipix2 chip, yields timing and charge measurements as well as precisespatial information in 3D of individual electron clusters. The Timepix ASIC sensitive areais arranged as a square matrix of 256 x 256 pixels of 55 x 55 µm2 size, resulting in a totaldetection area of ∼ 14× 14 mm2, which represents 87 % of the entire surface area. The chip isdesigned and manufactured in a six-metal 0.25 µm CMOS technology and contains ∼50 milliontransistors. Each pixel in the chip matrix can be programmed to record either the arrivaltime of the avalanche charge signal with respect to an external shutter (“TIME” mode) or the14-bit counter is incremented as long as the signal remains above the threshold (Time OverThreshold “TOT” mode), thus providing pulse-height information. The operation of MPGDwith Timepix chip has demonstrated the possibility to reconstruct 3D-space points of individualprimary electron clusters with ∼ 30 − 50 µm spatial resolution and event-time resolution withns precision (see Fig. 11b) 148,149,150. Thanks to these developments, a micro-pattern devicewith CMOS readout can serve as a high-precision “electronic bubble chambers” (see Fig. 11a).

a) b)

Figure 11: a) Image of the 5 GeV electron track recorded at the DESY testbeam with triple-GEM detector andTimePix CMOS ASIC operated in the mixed mode: every second pixel is operated in the “TOT”/”TIME” modesin the “chess board” fashion. The x, y-axis represent chip sensitive area, obtained by mapping the original data(matrix of 256 × 256 pixels of 55 µm pitch) into a 181 × 181 pixels matrix with a pitch of 78 µm. The color is themeasure of the time-over-threshold and drift time information, respectively; b) Spatial resolution as a function of

drift distance in Ar/CO2(70:30) mixture and triple-GEM/Medipix2 detector.

An elegant solution for the construction of the Micromegas with pixel readout is the integra-tion of the amplification grid and CMOS chip by means of an advanced “wafer post-processing”technology 33. With this technology, the structure of thin (1 µm) aluminum grid is fabricatedon top of an array of insulating (SU-8) pillars of typically 50 µm height, which stand above theCMOS chip, forming an integrated readout of the gaseous detector (see Fig. 12a). This tech-nology is called “InGrid” (developed by the collaboration of MESA+ institute of the Universityof Twente and NIKHEF) and allows an accurate alignment of hexagonal grid holes (∼ 60 µmpitch) with pixel input pads, sub-micron precision in the grid hole diameter and a constantthickness of the avalanche gap, which results in a uniform gas gain. The process uses standardphoto-lithography and wet etching techniques and is CMOS compatible. It can be used to equipboth single chips and chip wafers with Micromegas grid. Amongst the most critical items thatmay affect the long-term operation of “InGrid” concept is the appearance of breakdown acrossthe 50-100 µm amplification gap. This is a critical issue for Micromegas with CMOS readout,where a multiplication grid is placed directly above the naked Timepix chip, since for single stageamplification discharges in presence of heavy ionizing particles can not be completely eliminated.

One way to achieve protection is to cover the CMOS chip with a thin layer (3 to 20 µm)

16

of highly resistive amorphous silicon or silicon nitride (Si3N4) deposited directly on top of the

Timepix ASIC 151,152. This protective cover, called “SiProt”, intends to protect the chip fromhigh instantaneous spark currents, which could destroy the chip, and from the the evaporationof the Timepix metal input pads during sparks (see Fig. 12b). The performance of the “InGrid”detector with 4 µm “SiProt” layer is illustrated in Fig. 12c for 2 GeV electron beam at DESY.Due to its high sensitivity, “InGrid” detector can resolve single primary electrons 31,153. Thecolor is a measure of the electron arrival time in “TOT” mode, red color corresponds to theprimary ionization electrons produced close to the chip surface. The spread of the clustersin Fig. 12c indicates that the diffusion increases with the distance from the chip. The timerange in He/iC4H10 mixture is larger than in Ar/iC4H10, which indicates lower drift velocityin He-based mixture, while the primary ionization density (number of active pixels) is higher in

Ar-based mixture 154. The “InGrid” device with a narrow (1-1.5 mm) drift gap (“GOSSIP”)is currently proposed as one of the possible upgrade options for the ATLAS tracking system atsLHC, aiming for a spatial precision of about 20 µm 94.

a) b) c)

Figure 12: a) The Scanning Electron Microscopy (SEM) image of the “InGrid” detector: a Timepix chip, resistivesilicon-nitride (Si3N4) protection layer and insulating SU-8 pillars supporting a perforated Al grid. The pillarsare placed at the intersections of four adjacent pixels; the circular holes in the grid are centered onto the inputpad of each pixel; b) Photo of the “InGrid” structure with 7 µm Si3N4 protection layer; c) Images of 2 GeVelectron tracks recorded with “InGrid” detector in Ar/C4H10 and He/C4H10 mixtures in “TIME” mode; the

color is a measure of the arrival time of electrons.

One of the most exciting future applications of GEM and Micromegas devices with CMOSmulti-pixel readout could be position sensitive single photon detection. Recently, a UV photo-detector based on a semitransparent CsI photocathode followed by the fine-pitch GEM foil, thatmatches the pitch of a pixel ASIC (50 µm), has shown excellent imaging capabilities 155. Thephotoelectron emitted from CsI layer drifts into a single GEM hole and initiates an avalanche,which is then collected on the pixel CMOS analog chip. Due to the high granularity and largeS/N of the read-out system, the “center of gravity” of the single electron avalanche correspondsto the center of GEM hole. Accumulating thousands of such events produces “self-portrait”of the GEM amplification structure, shown in Fig. 13 (a) and allows to achieve superior sin-

gle electron avalanche reconstruction accuracy of 4 µm rms 155. Another monolithic gaseousUV-photon imaging device is based on the “InGrid” concept and reflective CsI photocathodedeposited on Micromegas grid. The photocathode deposition and its operation is found to beadequate in combination with both the “InGrid” detector and Timepix chip. Fig. 13b showsimage recorded with a steel mask placed in front of the “InGrid” detector under UV-photon ir-radiation. The detector operated reliably in He/iC4H10 mixture with single electron extractionefficiency of ∼ 50 % and a maximum achievable gain of ∼ 6× 104. These results encourage fur-ther developments towards high-resolution UV photon devices, based on combination of MPGDwith CMOS pixel readout. Futhermore, being made of UHV-compatible materials, “InGrid”concept could be used in cascaded visible light-sensitive imaging gas photomultipliers.

A key point that has to be solved to allow CMOS pixel readout of MPGD for applica-

17

a) b)

Figure 13: a) Cumulative map of hundreds of thousands events from the UV light conversion in the semitrans-parent CsI photocathode of the detector producing a kind of “self-portrait” of the GEM amplification structure.The white spots correspond to the individual GEM holes at 50 µm pitch; b) 2D image of the steel mask under

irradiation of the “InGrid” detector with reflective CsI photocathode in He/iC4H10 (80:20) mixture.

tions in various fields of fundamental science and beyond is the production of large area de-tectors. Presently prototypes under construction rely at maximum on 2 × 4 chips integratedreadout 97,157. Going to large surface detector requires solution for the dead space on one sideof the Timepix chip, where the electrical connections enter the chip (see Fig. 10b). A first solu-tion relies on etching of “through-silicon vias” on the Timepix chip that allows to bring out thesignals and services on the backside of the chip using “through-wafer-vias” technology 158,159.This is known as “via-last” operation: the existing Timepix wafers are modified by etching“in-vias” connections after the CMOS wafer production step is completed. In order to movefrom 3-side tileable detectors to 4-side, all common circuitry (bias, converters, ...) needs tobe spread over the chip area which is nowadays become possible thanks to the progress in 3Dmicro-electronics, where the different functions can be realized on different tiers and connectedthanks to “through-silicon vias” and inter-connects. However, a major R&D effort is requiredin the future to fully exploit this potential.

Properly integrated into large systems (including development of large area MPGD basedon the integrated multi-chip CMOS ASICs), the pixel readout may open new opportunities for

an advanced Compton Telescope 160,161, detection of Axions and Weakly Interacting MassiveParticles162, neutrino-less double beta decay experiments163 and 3D imaging of nuclear recoils.The primary advantage of a gaseous tracker with pixel readout is that the direction of theCompton electron or the low energy nuclear recoil can be reconstructed far more accuratelythan in any other detection medium.

8 Summary

Radiation detection and imaging with gas-avalanche detectors, capable of economically coveringlarge detection volumes with a low material budget, have been playing an important role in manyfields. While extensively employed at the LHC, RHIC, and other advanced HEP experiments,present gaseous detectors (wire-chambers, drift-tubes, resistive-plate chambers and others) havelimitations which may prevent their use in future experiments.

The possibility of producing micro-structured semi-conductor devices (with structure sizesof tens of microns) and corresponding highly integrated readout electronics led to the success ofsemi-conductor (in particular silicon) detectors to achieve unprecedented space-point resolution.Micro-pattern gas-amplification structures now open the possibility to apply the same technologyto gaseous detectors and enable a plethora of new detector concepts in science and industry,in medical and commercial applications. Microelectronics needs to keep up with the state-of-the-art developed by the microelectronics industry and the interconnection between electronicsand detectors needs to be improved to reduce the complexity and the material of interconnect

18

technologies.

The RD51 collaboration at CERN, approved in December 2008, aims at facilitating thedevelopment of advanced gas-avalanche detector technologies and associated electronic-readoutsystems, for applications in basic and applied research. The main objective of the R&D programis to advance technological development and application of micro-pattern gaseous detectors. TheRD51 collaboration involves ∼ 430 authors, 73 Universities and Research Laboratories from 25countries in Europe, America, Asia and Africa. All partners are already actively pursuing eitherbasic- or application-oriented R&D involving a variety of MPGD concepts. The collaborationestablished common goals, like experimental and simulation tools, characterization concepts andmethods, common infrastructures at test beams and irradiation facilities, and infrastructure forMPGD production. More information can be found at the RD51 collaboration webpage 34.

References

1. H. Geiger, E. Rutherford, Proc. Royal Soc. A 81, 141 (1908).2. H. Geiger, W. Mueller, Phys. Zeits. 29, 839 (1928).3. G. Charpak et al, Nucl. Instrum. Methods A 62, 262 (1968).4. F. Sauli, “Principles of Operation of Multiwire Proportional and Drift Chambers”, CERN

77-09 (1977).5. G. Charpak, F. Sauli, Ann. Rev. Nucl. Part. Sci. 34, 285 (1984).6. F. Sauli, “From Bubble Chambers to Electronic Systems: 25 Years of Evolution in Particle

Detectors at CERN”, CERN PH-EP 2004-040 (2004).7. C. Grupen, Particle Detectors (Cambridge Univ. Press, 1996).8. W. Blum, W. Riegler, L. Rolandi, Particle Detection with Drift Chambers (Springer-Verlag,

2008).9. A. H. Walenta et al, Nucl. Instrum. Methods A 92, 373 (1971).

10. R. Bouclier et al, Nucl. Instrum. Methods A 265, 78 (1988).11. H. Drumm et al, Nucl. Instrum. Methods A 176, 333 (1980).12. D.R. Nygren, J.N. Marx, Phys. Today, 31, Vol. 10 (1978).13. A. H. Walenta, IEEE Trans. Nucl. Sci. (TNS), Vol. 26, 73 (1979).14. G. Charpak, F. Sauli, Phys. Lett. B 78, 523 (1978).15. J. Seguinot, T. Ypsilantis, Nucl. Instrum. Methods A 142, 377 (1977).16. A. Oed, Nucl. Instrum. Methods A 263, 351 (1988).17. A. Barr et al, Nucl. Phys. B (Proc. Suppl.) 61B, 264 (1998).18. A. Oed, Nucl. Instrum. Methods A 490, 223 (2002).19. F. Sauli, Nucl. Instrum. Methods A 386, 531 (1997).20. Y. Giomataris et al, Nucl. Instrum. Methods A 376, 29 (1996).21. F. Sauli, A. Sharma, Ann. Rev. Nucl. Part. Sci. 49, 341 (1999).22. M. Titov, Nucl. Instrum. Methods A 581, 25 (2007).23. M. Titov et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 49(4), 1609 (2002).24. M. Titov, “Radiation Damage and Long-Term Aging in Gas Detectors,”, arXiv:

physics/0403055; Proc. of the Workshop of the INFN ELOISATRON Project, “InnovativeDetectors For Super-Colliders,” Erice, Italy, Sept.28 - Oct.4 (2003).

25. L. Periale et al, Nucl. Instrum. Methods A 478, 377 (2002).26. R. Chechik et al, Nucl. Instrum. Methods A 535, 303 (2004).27. A. Breskin et al, Nucl. Instrum. Methods A 598, 107 (2009).28. A. Di Mauro et al, Nucl. Instrum. Methods A 581, 225 (2007).29. E. Costa et al, Nature 411, 662 (2001).30. R. Bellazzini et al, Nucl. Instrum. Methods A 535, 477 (2004).31. M. Campbell et al, Nucl. Instrum. Methods A 540, 295 (2005).

19

32. A. Bamberger et al, Nucl. Instrum. Methods A 573, 361 (2007).33. M. Chefdeville et al, Nucl. Instrum. Methods A 556, 490 (2006).34. http://rd51-public.web.cern.ch/RD51-Public

35. “Review of Particle Physics” (By Particle Data Group), Chapters 28.7.1-28.7.3

‘‘Gaseous Detectors’’ written by F. Sauli, M. Titov, Phys. Lett. B 667, 292 (2008).36. http://rjd.web.cern.ch/rjd/cgi-bin/cross

37. A. Peisert, F. Sauli, “Drift and Diffusion of Electrons in Gases”, CERN 84-08 (1984).38. S. Biagi, Nucl. Instrum. Methods A 421, 234 (1999).39. http://consult.cern.ch/writeup/magboltz/

40. L.G. Christophorou et al, “Electron Interaction in CF4”, J. Phys. Chem. Ref. DataVol.25, No.5, 1341 (1996).

41. E. McDaniel, E. Mason, The Mobility and Diffusion of Ions in Gases (Wiley, 1973).42. G. Shultz et al, Rev. Phys. Appl. 12, 67 (1977).43. G. Charpak et al, Nucl. Instrum. Methods A 167, 455 (1979).44. A. Breskin et al, Nucl. Instrum. Methods A 124, 189 (1975).45. A. Breskin et al, Nucl. Instrum. Methods A 156, 147 (1978).46. D. Nygren et al, PEP-PROPOSAL-004, Appendix A 6, (1976).47. D. Decamp et al, Nucl. Instrum. Methods A 294, 121 (1990).48. Y. Sacquin et al, Nucl. Instrum. Methods A 323, 209 (1992).49. S. Afanasev et al, Nucl. Instrum. Methods A 430, 210 (1999).50. H. Wieman et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 44(3), 671 (1997).51. http://www.ansoft.com

52. D. Antonczyk et al, Nucl. Instrum. Methods A 565, 551 (2006).53. T. Meyer, “System Aspects of (Gaseous) Tracking Detectors”, Proc. of the Workshop of

the INFN ELOISATRON Project, “Innovative Detectors For Super-Colliders,” Erice, Italy,Sept.28 - Oct.4 (2003).

54. M. Aleksa et al, Nucl. Instrum. Methods A 446, 435 (2000).55. R. Bouclier et al, Nucl. Instrum. Methods A 381, 289 (1996).56. V.Peskov et al, Nucl. Instrum. Methods A 397, 243 (1997).57. Yu. Ivaniouchenkov et al, Nucl. Instrum. Methods A 422, 300 (1999).58. A. Bressan et al, Nucl. Instrum. Methods A 424, 321 (1999).59. Yu. Ivaniouchenkov et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 45(3), 258 (1998).60. P.Fonte et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 46(3), 321 (1999).61. C. Iacobaeus et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 49(4), 1622 (2002).62. V. Peskov, P. Fonte, “Research on discharges in micropattern and small gap detectors”,

arXiv: 0911.0463(2009).63. A. Bondar et al, Nucl. Instrum. Methods A 556, 495 (2006).64. A. Bressan et al, Nucl. Instrum. Methods A 425, 262 (1999).65. S. Bachmann et al, Nucl. Instrum. Methods A 438, 376 (1999).66. S. Bachmann et al, Nucl. Instrum. Methods A 470, 548 (2001).67. S. Bachmann et al, Nucl. Instrum. Methods A 479, 294 (2002).68. A. Breskin et al, Nucl. Instrum. Methods A 513, 250 (2003).69. J.F.C.A. Veloso et al, Rev. Sci. Instr. 71(6), 2371 (2000).70. A. Lyashenko et al, JINST 2, P08004 (2007).71. Y. Giomataris, Nucl. Instrum. Methods A 419, 239 (1998).72. J. Derre et al, Nucl. Instrum. Methods A 459, 523 (2001).73. G. Charpak et al, Nucl. Instrum. Methods A 478, 26 (2002).74. J. Derre et al, Nucl. Instrum. Methods A 449, 314 (2000).75. Y. Giomataris et al, Nucl. Instrum. Methods A 560, 405 (2006).76. S. Anvar et al, Nucl. Instrum. Methods A 602, 415 (2009).

20

77. S. Andriamonje et al, JINST 5, P02001 (2010).78. L. Periale et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 50(4), 809 (2003).79. C. Shalem et al, Nucl. Instrum. Methods A 558, 475 (2006).80. R. Chechik et al, Nucl. Instrum. Methods A 553, 35 (2005).81. C. Azevedo et al, JINST 5, P01002 (2010).82. R. Oliveira et al, Nucl. Instrum. Methods A 576, 362 (2007).83. A. Di Mauro et al, Nucl. Instrum. Methods A 581, 225 (2007).84. A. Di Mauro et al, Nucl. Instrum. Methods A 610, 169 (2009).85. B. Ketzer et al, Nucl. Instrum. Methods A 535, 314 (2004).86. C. Bernet et al, Nucl. Instrum. Methods A 536, 61 (2005).87. D. Neyret et al, JINST 4, P12004 (2009).88. M. Alfonsi et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 51(5), 2135 (2004).89. S. Lami, “First Test Results for the TOTEM T2 Telescope”, 2009 IEEE NSS Conference

Record.90. M. Vandenbroucke, “A GEM-based TPC Prototype for PANDA”, 2009 IEEE NSS Con-

ference Record.91. http://hallaweb.jlab.org/12GeV/SuperBigBite/

92. F. Simon et al, “The Forward GEM Tracker of STAR at RHIC”, 2009 IEEE NSS Confer-ence Record.

93. G. Barouch et al, Nucl. Instrum. Methods A 423, 32 (1999).94. H. van der Graaf et al, “Performance and prospects of GridPix and Gossip detectors”,

RD51 Note, RD51-2009-006, (2009).95. A. Balla et al, “Status of the Cylindrical GEM Project for the KLOE-2 Inner Tracker”,

2009 IEEE NSS Conference Record.96. A. Aune et al, Nucl. Instrum. Methods A 604, 53 (2009).97. G. De Lentdecker, “A Large TPC Prototype for an ILC Detector”, 2009 IEEE NSS Con-

ference Record.98. D.C. Arogancia et al, Nucl. Instrum. Methods A 602, 403 (2009).99. T. Matsuda, JINST 5, P01010 (2010).100. Jaehoon Yu, “Development of GEM Based Digital Hadron Calorimeter”, 2009 IEEE NSS

Conference Record.101. C. Adloff et al, JINST 4, P11023 (2009).102. R. Chechik et al, Nucl. Instrum. Methods A 502, 195 (2003).103. R. Chechik, A. Breskin, Nucl. Instrum. Methods A 595, 116 (2008).104. A. Buzulutskov, Phys. Part. Nucl. 39, 424 (2008).105. Y. Giomataris, G. Charpak, Nucl. Instrum. Methods A 310, 585 (1991).106. Z. Fraenkel et al, Nucl. Instrum. Methods A 546, 466 (2005).107. C. Woody et al, “Initial Performance of the PHENIX Hadron Blind Detector at RHIC”,

2009 IEEE NSS Conference Record.108. A. Di Mauro et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 56(3), 1550 (2009).109. L. Periale et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 52(4), 927 (2005).110. L. Periale et al, Nucl. Instrum. Methods A 573, 302 (2007).111. A. Bondar et al, Nucl. Instrum. Methods A 581, 241 (2007).112. A. Bondar et al, Nucl. Instrum. Methods A 598, 121 (2009).113. C. Hagmann et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 51(5), 2151 (2004).114. A. Rubbia, “ Experiments for CP violation: A Giant liquid argon scintillation, Cerenkov

and charge imaging experiment?” arXiv: hep-ph/0402110 (2004).115. A. Rubbia, “ ArDM: A Ton-scale liquid Argon experiment for direct detection of dark

matter in the universe,” arXiv: hep-ph/0510320 (2005).116. A. Buzulutskov, Instr. Exp. Tech. 50, 287 (2007).

21

117. C. Grignon et al, Nucl. Instrum. Methods A 571, 142 (2007).118. S. Duval et al, JINST 4, P12008 (2009).119. G. Smith, “Gas-based detectors for synchrotron radiation”, Journal of Synchrotron Ra-

diation 13(2), 172 (2006).120. D. Pacella et al, JINST 1, P09001 (2006).121. S. Duarte Pinto et al, “Spherical GEMs for parallax-free detectors”, 2009 IEEE NSS

Conference Record.122. S. Aune et al, Nucl. Instrum. Methods A 604, 15 (2009).123. J. Pancin et al, Nucl. Instrum. Methods A 524, 102 (2004).124. M. Houry et al, Nucl. Instrum. Methods A 557, 648 (2006).125. S. Andriamonje et al, 2004 IEEE NSS/MIC Conference Record.126. S. Andriamonje et al, Nucl. Instrum. Methods A 562, 755 (2006).127. C. Iacobaeus et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 48, 1496 (2001).128. J. Ostling et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 50, 809 (2003).129. S. Fetal et al, Nucl. Instrum. Methods A 513, 42 (2003).130. M. Hohlmann et al, “Design and Construction of a First Prototype Muon Tomography

System with GEM Detectors for the Detection of Nuclear Contraband,” arXiv: 0911.3203(2009).

131. G. Charpak et al, “ Performance of wire-type Rn detectors operated with gas gain inambient air in view of its possible application to early earthquake predictions,” arXiv:1002.4732 (2010).

132. G. Charpak et al, JINST 4, P12007 (2009).133. G. Charpak et al, IEEE Trans. Nucl. Sci. (TNS), Vol. 55, 1657 (2003).134. S. Duarte Pinto et al, JINST 4, P12009 (2009).135. T. Alexopoulos et al, JINST 4, P12015 (2009).136. F. Tessarotto, Talk at the 3rd RD51 Collaboration Meeting, June 12-17 (2009), Crete,

Greece.137. A. Lyashenko et al, JINST 4, P07005 (2009).138. G. Charpak et al, JINST 4, P12007 (2009).139. R. Bellazzini et al, “A Micro Pattern Gas Detector for X-ray polarimetry”, Proc. SPIE,

Polarimetry in Astronomy, 4843, 372 (2003).140. R. Bellazzini et al, “The sensitivity of a photoelectric X-ray polarimetry for astronomy:

the impact of gas mixtures and pressure”, Proc. SPIE, Polarimetry in Astronomy, 4843,394 (2003).

141. F. Muleri et al, Nucl. Instrum. Methods A 584, 149 (2008).142. R. Bellazzini et al, Nucl. Instrum. Methods A 560, 425 (2006).143. R. Bellazzini et al, Nucl. Instrum. Methods A 566, 552 (2006).144. R. Bellazzini et al, Nucl. Instrum. Methods A 572, 160 (2006).145. X. Llopart, M. Campbell, IEEE Trans. Nucl. Sci. (TNS), Vol. 49(5), 2279 (2002).146. X. Llopart et al, Nucl. Instrum. Methods A 581, 485 (2007).147. P. Colas et al, Nucl. Instrum. Methods A 535, 506 (2004).148. A. Bamberger et al, Nucl. Instrum. Methods A 581, 274 (2007).149. A. Bamberger et al, arXiv: 0709.2837, (2007).150. C. Brezina et al, JINST 4, P11015 (2009).151. A. A. Aarts et al, “Discharge Protection and Ageing of Micromegas Pixel Detectors”,

2006 IEEE NSS Conference Record.152. M. Bosma et al, “Results from MPGDs with a Protected Timepix or Medipix2 pixel

sensor as active anode”, 2007 IEEE NSS Conference Record.153. V. M. Blanco Carballo et al, Nucl. Instrum. Methods A 583, 42 (2007).154. Y. Bilevych et al, “The Performance of GridPix Detectors”, 2009 IEEE NSS Conference

22

Record.155. R. Bellazzini et al, Nucl. Instrum. Methods A 581, 246 (2007).156. J. Melai et al, arXiv: 1003.2083, (2010).157. J. Kaminski, Talk at the 4th RD51 Collaboration Meeting, Nov. 23-25 (2009).158. K. Takahashi et al, Microelectronics Realibility4312672003.159. E. Heijne, Nucl. Instrum. Methods A 541, 274 (2005).160. A. Takada et al, Nucl. Instrum. Methods A 546, 258 (2005).161. K. Miuchi et al, Nucl. Instrum. Methods A 576, 43 (2007).162. S. Ahlen et al, Intl. Jour. Mod. Phys. A, 25(1), 1 (2010).163. V. Blanco Carballo et al, JINST 5, P02002 (2010).

23